Plywood Subfloor Seams

[cx]

Especially one without a vapor retarder over a properly graded surface and good ventilation or partial conditioning.

[T_Hulse]

The most you can get with a crawlspace (standing exposed water & vents closed off) is a moisture percentage in the low 20's for your subfloor, which is still less than the full-blown rain exposure that is common during framing (especially here! ).
Even if it were somehow higher than framing, it still wouldn't cause edge peaking; that is not from plain horizontal expansion of sheets. Peaking is from a differential in moisture caused by different amounts of access that water has to different areas of the board, not just by "lots" of moisture; and also by the super-absorbent end grain that sucks up liquid water (not just moisture) faster than the surface. Hardwood floors for instance. We've all seen the peaking when they get wet, but no matter how wet they got, they would never peak if they didn't have a waterproof coating on top and had tile on top that didn't allow the top to air dry much faster than the bottom. As the bottom gets wet & expands, the dry top becomes relatively shorter and causes the whole board to curl. So moisture differential & absorbent end grain, it has nothing to do with compression at the joints. Horizontal expansion doesn't cause edge peaking.
Edit: to clarify, also "moisture" by itself can't cause edge peaking. It has to be liquid, running water or a differential in moisture which is not present in a finished tile assembly.

[dhagin]

I've built a few homes in the PNW, and don't necessarily disagree with any of the arguments the "no gap needed" crowd makes. This assumes, of course, that the maximum moisture those panels see is when they're initially installed.

Would anyone gap any sheets, subfloor or underlay, if the project was framed and dried-in before the panels got wet? I've been concerned about this one especially, as we'll occasionally have a few relatively warm & very dry months followed by a year of relatively cool & wet ones. Further complicating issues is folks who leave town in the winters here, during the cool & wet months, turning their heat way down below what I suggest.

[T_Hulse]

Dana, that's another whole separate reason this gap thing doesn't work. That gap in the subfloor is not for expansion, it's for glue.
A third reason is that when you say gap, you really mean "expansion joint"; and an expansion joint in the construction that is not carried through to the tile surface is a violation and a waste of time. You dont' want to encourage the subfloor panels to have any more independent movement (from each other) than they already have. It would be dangerous and cause tile cracking if it actually worked as an expansion joint.

[dhagin]

Originally Posted by Tom

That gap in the subfloor is not for expansion, it's for glue.

What's your source Tom?

These gaps I'm referring to above are not expansion or movement joints as described by EJ171.

[T_Hulse]

There's one source posted in this thread yesterday, the Advantech instructions that wants polyurethane adhesive between the tongue and groove. They show a picture explaining the reason you don't use too much glue is that you'll have to scrape it off the top (not that somehow the boards could move independently if you used less glue). When you built those houses, did you have glue in the T & G?
When you say gap, if you are thinking that somehow that gap was for expansion & contraction, well that's just other words for movement. You are describing a construction joint, like an EJ171A wood version, and if construction joints are designed for movement (whether in concrete, wood, or other), then they must be honored all the way through to the tile surface. The reason there is no EJ171A for wood, like you are describing, is that wood floors are not designed to intentionally encourage independent movement between sheets.
An expansion joint that is not carried through to the tile surface is a worthless expansion joint (and potentially harmful if it actually works).

[John Bridge]

I believe the 1/8 in. required gap IS for expansion during the course of construction.

[Lump]

12 percent max moisture content is also a good rule of thumb to go by prior to installing any flooring....even backer board. Take the time and protect the gaps.

[dhagin]

I guess i don't get it.

TCNA F143-12 Plywood Subfloor Plywood Underlayment Ceramic Tile, for instance, requires gaps (their term) between sheets of both subfloor & underlayment. These gaps are to be offset minimum 2" so naturally don't carry through the assembly.

APA also recommends gaps or spaces between sheets of subfloor & underlayment, . Their stated reason is to allow for expansion.

a few APA ref's:
L335.pdf

M300.pdf

[eurob]

Dana, From the first APA link , is this gap -- spacing of 1/32 recommended at underlayment butt joints -- or space that you are refering to ?
Page 4 and 5 from the same link -- L335 -- are pretty interesting .

[Lump]

Dana, good link on the L335pdf

[dhagin]

Yes Roberto on the 1/32. TCNA says 1/8, just to give us something to discuss.

Here's a few other docs referenced in those others for tile.

TT-006.pdf

Form E30 is too big to upload. Go here and you can get it yourself from APA, it's on the top of the right column. Probably have to register, which is free & instant.

http://www.apawood.org/searchResults...x=0&submit.y=0

[eurob]

The 1/16 and 1/8 for the gap is for the T&G applications from the other link -- M300.

TCNA is in the safe zone -- 1/8 gap on all installations --

Thanks Dana for the links . and thanks to Tom -- T_Hulse -- for taking the time to explain it .

[T_Hulse]

Dana, F143 does describe subfloor gaps, but that does not conflict with what I've been saying, as in the Advantech description of them getting filled with glue, like your framers probably did on the houses you built. F143 is not intended to conflict with manufacturer's directions explaining those gaps get glue.
F143 also describes 1/4" gaps in plywood underlayments, but they take the neutral way out and don't comment on filling or not filling with thinset, leaving it to setting material recommendations. There's disagreement among the manufacturers, but some of them like Custom have it dead wrong; not understanding that seam peaking can not be caused by expansion, it takes liquid water.
The only time you would have to carry them all the way through the tile assembly is in the odd case if you insisted they must be kept open with no hard glue or thinset, or that they needed flexible caulk in them or tape over them. Those conditions all say you are intending to use them as moving construction joints, which are inappropriate for a wood floor, but would require a full soft joint over them if you really think they work.
Also, L335 from the APA that you mentioned does allow for 1/32" gaps for expansion in underlayments, like you said. But keep reading. They also say those gaps should be filled before the floor covering is installed.

[dhagin]

That framer of which you speak would be me. And i can assure you that i've never glued a t&g seam, seen a spec to glue plywood subfloor edges or t&g, or installed advantec to my knowledge. I've also built homes without t&g and blocked everything.

Saw the bit about filling underlayment gaps before installing the finish floor, but believe that to be in reference to resiliant floors, not tile, hardwood, carpet, etc... but not entirely sure.

I certainly respect your opinions Tom and believe i understand the argument, but i come at this from a builders background having more than a couple a decades of putting houses together and just can't wrap my brain around *not* gapping plywood on all sides.